CrossFit Marketing Automation Checklist: The Two Weeks After a Free Trial

CrossFit marketing automation best practices for the two weeks after a free class come down to one split. Automate the reminders, the same-day thank-you, booking nudges on days two and three, the membership offer and a day-twelve last call. Keep day-one feedback, injury questions, price conversations and the invitation to join personal from the coach.

Automation should stop the moment a person books, buys, replies or opts out. That one rule removes most of the robotic feeling, and it is the first thing to build into whatever software you use. Everything else is timing. The two weeks after a trial are when a visitor weighs up whether your box is for them, and what they hear from you in that window is either easy to act on or noise.

Follow me on Instagram@sagnikteaches

CrossFit marketing automation best practices, day by day

Every item is marked automated or personal, with the reason and a quick way to check it worked. Use it as a build list the first time and as a weekly spot-check afterwards.

Connect on LinkedInSagnik Bhattacharya

Day 0: from booking to the same-day thank-you

• Automated: booking confirmation. Sent at booking with the time, where to enter, what to wear, the coach's first name and the waiver link. It also invites them to reply with any injury or worry, which routes to the coach. Check: book a test trial from your own phone and read every line as a stranger would.
• Automated: reminders 24 hours and 2 hours before. Short, with the address and a reply route. Check: a same-day booking skips the 24-hour reminder rather than sending it late.
• Personal, 30 seconds: two notes straight after class. One specific thing they did well, one thing to ask about, such as a goal or a worry. These notes feed every personal message that follows. Check: every trial has both notes logged by 8 pm.
• Automated: same-day thank-you within two hours. A text or email from the box, not from a no-reply address, saying thanks and that a reply reaches the coach. Check: sample timestamps, and make sure nothing goes out during quiet hours.

Days 1 to 3: the personal window

• Personal: the coach's message on day one. Three sentences that mention something the person actually did, with one easy question at the end. AI may draft it from the two notes; the coach edits and sends it. Check: sent by midday, nothing medical in writing, logged against the lead.
• Automated: day-two email on how to book a second class. Beginner-friendly class times, class size, how booking works, and what the on-ramp is if you run one. Check: open the link on a phone; it should show only classes that still have places.
• Automated, conditional: day-three nudge. Only for people with no second class booked: "Want me to hold a spot for Thursday at 6?" A yes reply goes straight to the coach. Check: a test lead who booked last night does not receive it.

Days 4 to 7: make the second class easy

• Automated: day-five email on what the first month looks like. The programme, the weekly rhythm and what a typical newcomer does. No promises about results. Check: read it for any sentence that predicts how someone will look or feel.
• Automated: day-seven membership options with a real end date. Two or three options, prices from your current list, and an offer that really expires on day 14. Check: the prices match the list and the end date is set in the tool, not typed from memory.
• Personal: a 15-minute chat invitation for people who clicked pricing and have not booked. Treat clicks as a hint, not proof: Mailchimp's own guidance on bot clicks says security tools can register as clicks and inflate open and click rates. Check: the coach has a short list each morning.

Days 8 to 12: the decision

• Automated: day-eight email answering five questions. Price, schedule, being the weakest in the room, injuries and contract length. Each answer matches your real policy. Check: have a coach read it and mark every answer they would not give out loud.
• Personal: day-ten text or call for people who have not booked. Two sentences and no pressure, from the coach's own number if you can. Check: outcome logged within the hour: booked, not now, or stop.
• Automated: day-twelve last call. The offer ends in 48 hours, and the message says what happens after: nothing bad, just the standard price. Check: the end date is the same one in the day-seven email.

Days 13 and 14: close the loop

• Automated or personal: the close. A polite message saying you will stop here, with an option for occasional updates. Personal is better for anyone who engaged and did not buy. Check: after day 14 nobody remains in the trial sequence.
• Personal, five minutes: update the lead's status. Booked, member, not now, or lost, and why in one line. Check: your weekly count of trials, bookings and members adds up to the trials you ran.

For how this sits inside your software, CrossFit marketing automation software compared for free-trial leads covers tools, and what happens after someone joins, from on-ramp to win-back, is in CrossFit on-ramp courses and win-backs. The list above is deliberately software-neutral.

Subscribe on YouTube@sagnikteaches

Stop rules: when every automated message must halt

Most complaints about automated follow-up come from sequences that carry on after something has changed. Build these stops first and test each one before you switch anything on.

The reply rule matters most. A person who writes "can I come Thursday?" and then receives a template saying "have you booked yet?" the next morning has just learned that nobody is reading. If your software cannot pause on a reply, find that out before you launch, not after.

What stays a personal message from the coach

The line between automated and personal is judgement. A system can send a time and a link. It cannot decide how to answer someone who says they felt out of their depth. These stay with a person:

• Day-one feedback. It is the first message from a human, and the one people remember.
• Anything about injuries, pain or health. The coach decides whether to raise it, and usually in person, not in writing.
• Price conversations and exceptions. A hardship request or a "can I pay monthly?" is a decision, not a template.
• Every reply to a question. AI can suggest wording; a person sends it.
• The invitation to join. The offer email is automated; the personal "I think you'd like the Tuesday group" is not.
• Anyone who says no. A kind, short acknowledgement from a person keeps the door open far better than a fifth automated nudge.

The coach's cost is about two minutes per trial. Make it smaller with a fixed note format: after class, two lines. Here is a made-up pair of notes and a prompt that turns them into a draft.

Notes after class:
Did well: stayed with the group on the rower, asked good questions about scaling.
Ask about: mentioned an old shoulder injury; wants to feel fitter for weekend hikes.

Write a three-sentence message from [coach first name] to [first name] the day after a free trial class. Use only these notes. Warm and specific, no hype. Do not mention any injury or health issue in writing. End with one easy question and offer a quick chat at [day and time].

An illustrative reply:

Hi [first name], thanks for coming in yesterday. You kept up on the rower and asked exactly the right questions about scaling, which is what a first class is for. Anything you want to ask before your next one? I'm around at [day and time] if you'd like a quick chat. [coach first name]

Two things for the coach to fix. The draft dropped the hiking goal, which is the most personal detail in the notes and not a health matter: adding "I'd love to help you feel fitter for those weekend hikes" turns a pleasant note into one that sounds like it was written for them. And the shoulder: the prompt kept it out of writing on purpose, so the coach should raise it at the next class, face to face, where they can ask what the person is comfortable doing. Without that rule, assistants often write "how's the shoulder?", which is well meant and exactly the sort of line that should not arrive by text.

Five messages written out, three of them automated

These are shapes to adapt, not wording to paste. Placeholders show where your own details go.

Day 0, automated text. Hi [first name], thanks for coming to [box name] today. Got a question about the class or booking another? Reply here and [coach first name] will answer. Reply STOP to opt out.

Day 1, personal. The coach's message above, edited so it sounds like the coach.

Day 3, automated and conditional. Hi [first name], want me to hold a spot for [day] at [time]? Reply YES and I'll book you in. [coach first name], [box name]

Day 7, automated email. Subject: Your options after the first class. Body: Thanks again for coming in. If you'd like to keep going, here are the three options, with prices as on our current list: [option 1], [option 2], [option 3]. The introductory offer runs until [date]. Questions? Reply and a coach will answer.

Day 14, close. Hi [first name], I'll stop messaging after this one. If now isn't the right time, no problem at all. If you'd like occasional updates about open classes and events, reply YES. Either way, thanks for trying us. [coach first name]

Two checks apply to all five. First, the "I" in the day-three message is the coach, so replies must reach the coach; an automated message that pretends to be personal but hits a dead inbox is worse than an honest one signed by the box. Second, the day-seven offer must expire on the date it states. A deadline that quietly resets for the next person teaches your leads that your deadlines are not real.

Where AI helps in this sequence, and where it stays out

AI earns its place in four jobs here. It writes variants of each message so you can test two wordings. It turns the coach's notes into drafts. It sorts the replies that arrive. And it summarises the month's trials for the owner. It stays out of sending replies by itself, of anything about injuries or health, of pricing exceptions, of promises about results, and of deciding who is "serious" and who is not.

Sorting replies is the least obvious use and the most useful when a coach is busy. Paste a batch of replies, with names removed, and ask for categories.

Sort each reply below into one category: Ready to book; Question about price or schedule; Worry or injury; Not now; Stop.
Return a table with the reply text, the category and, for Ready to book and Question only, the first three words of a suggested response.
Do not draft anything for Worry or injury, and do not draft anything for Stop.

On five made-up replies, an illustrative result looks like this:

Check the sort against your own reading for the first fortnight. Models are good at the obvious cases and shaky on short, ambiguous replies such as "maybe" or a thumbs-up. The rule is that a person reads every Worry or injury reply, and that "stop" always lands in the opt-out list, whatever the model thought it meant. For how AI handles first enquiries before the trial, see can AI answer gym enquiries and book trial sessions.

The numbers to read at day 14

Four counts tell you whether the sequence works. Use a month with 30 trial classes. These readings are illustrative, not benchmarks, and your own before-and-after comparison matters more than any average.

Read them as a set. If the second-class figure rises but day-14 membership stays flat, the offer or the price conversation needs attention. If membership rises but opt-outs climb, the sequence is pushing too hard: remove a message. If replies go unanswered, no amount of automation upstream will help, because the human step is the bottleneck. Compare a month before the sequence with a month after, and change one thing at a time.

Consent, quiet hours and opt-outs: set these before switching on

• Consent. A separate, unticked box on the booking form that names text messages, with the answer and timestamp stored against the lead.
• Identification. The first text names the box, so nobody wonders who is writing.
• Opt-out. A simple word such as STOP that works on every channel and takes effect immediately, plus an unsubscribe link in every email.
• Quiet hours. No sends at night. Gymdesk's page on texting, for example, offers quiet hours so nothing fires at 2 am, and says opt-outs are handled for you. Check that yours does the same.
• Cost. Texts are often metered. The same Gymdesk page describes SMS as a usage-based add-on, a small monthly fee for the texting number plus a per-message rate, with email and push included in the plan. Thirty trials with three texts each is 90 texts a month; multiply by your rate.
• Rules. They differ between places, so ask an adviser what applies to you.

One lead, fourteen days: the sequence in action

A walk-through shows how the stop rules cut the sequence short. Follow a newcomer who books a free Saturday 9 am class on a Wednesday. Everything here is illustrative.

1. Wednesday. The confirmation and waiver link arrive within a minute.
2. Friday and Saturday. The 24-hour reminder lands on Friday morning and the 2-hour reminder at 7 am on Saturday.
3. Saturday. The class runs. By 10:15 the coach has logged two notes. At 11:00 the thank-you text goes out, signed by the box.
4. Sunday. The coach sends the day-one message at 10:30, built from the notes and edited.
5. Monday. The day-two email arrives. The lead clicks the booking link but does not book.
6. Tuesday. The day-three nudge asks whether to hold a spot on Thursday at 6. The lead replies "Yes please." The system pauses, assigns the lead to the coach, who books the class, and the lead moves to the pre-member welcome.

The day-five, day-seven, day-eight and day-twelve messages never send. That is the system working: this person needed two good touches, not six. A second lead stays quiet. The day-five, day-seven and day-eight emails go out, and on day ten the coach texts: "No pressure at all. Is there anything that would make it easier to come back?" The reply: "Not right now, money is tight." The coach answers kindly, offers to check in after a month, marks the lead Not now with the reason "price", and the system stops everything. There is no day-twelve last call and no day-fourteen close, because a person decided they were no longer needed.

A thirty-trial month: what the checklist costs in time

At 30 trials a month, a six-touch follow-up by hand at three minutes a touch is 540 minutes, or nine hours. After automation, the coach still does the personal parts, and the system does the rest.

That is a saving of about six hours a month, against roughly four hours to write the messages, build the flow and test it once. On these assumptions the set-up time is recovered inside the first month. More important, the hours that remain are the ones a coach should be spending anyway. A tool can handle the rest of what a studio has to say before and after; fitness studio emails for welcome, waitlist and lapsed members shows the same split in a studio setting.

Mistakes that make a sequence feel like a bot

• The sequence continues after a reply. Test the pause rule with a dummy lead before launch.
• Messages about a class they did not attend. No-shows get their own path.
• Automated messages posing as personal ones. "It's [coach first name] here" when no human wrote it is a small lie, and a lead who replies to a dead inbox concludes the whole box is careless.
• The same wording in every email. Write a different angle for each message, and use AI to draft alternatives rather than copy one template five times.
• A deadline that resets. An offer that expires "in 48 hours" every time someone reads it is a false deadline.
• Health talk in a text. Injuries are for a conversation.
• Late sends. A message at 10 pm feels like pressure whatever it says. Use quiet hours.
• A trial lead on the member list. Keep trial and member communications separate, or a trial gets the class-cancellation notice meant for someone else. A good reminder routine for members is covered in how to reduce no-shows with AI reminders and automatic rebooking.

Questions box owners ask about trial follow-up

How many automated messages are too many in two weeks?

There is no fixed number, but about six automated touches and three personal ones in 14 days is a workable ceiling for a first sequence. Stop the sequence the moment the person books, buys, replies or opts out. If opt-outs climb above your usual level, remove a message rather than tweaking its wording.

Should the day-one message after a trial be automated?

Automate the same-day thank-you, but keep day one personal. A coach's note that mentions something the person actually did lands differently from a template. Have the coach jot two notes after class, let AI turn them into a three-line draft, and have the coach edit and send it. Two minutes per trial is realistic.

Should I discount the first month to convert trials?

Test it rather than assume it. A discount lowers the price every later month is judged against. Try waiving an on-ramp or joining fee, or a free guest pass, and compare conversion at day 14. Keep any offer short and dated, with a genuine end date, and print the standard price beside it.

What if the person did not show up to the free class?

Treat it as a separate path. Send one message within the hour offering another slot, one the next day, and then stop. A no-show has not had the experience you would be following up, so the sales sequence does not apply. Let the coach phone anyone who rebooked twice and missed both.

Do I need consent to text a trial lead?

Get it on the booking form with a separate, unticked box that names texts, record it with a timestamp, identify the box in the first message, include a simple opt-out and send only in sensible daytime hours. Rules differ between places, so ask an adviser what applies to you before you switch texts on.

Further reads

• Gym Automation Mistakes That Increase Member Churn — Automation errors that push members out the door
• BJJ Gym Marketing: Trial-Class Follow-Ups and Belt Posts With AI — The same trial problem in a martial arts gym
• How Pilates Studios Use AI to Turn Intro Offers Into Memberships — Turn intro offers into memberships in a studio
• Gym Referral Programme: Set Rewards and Test the Maths With AI — Reward referrals without losing money
• Can AI Answer Trial-Class Calls for a Martial Arts School? — Handle trial-class phone calls with AI
• Can AI Write CrossFit Programming? Scaling Options and Coach Notes — Where AI helps with scaling notes for coaches
• AI Automation for Gyms: Failed Payments, Freezes and Win-Back Texts — Automate failed payments and freezes after they join
• Automate Pilates Fundamentals Tips: A New-Client Email Series With AI — A five-email pre-class series triggered by booking: AI drafts from your instructor's cue notes, the instructor approves each email, then it sends itself.
• AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.

Sources: Gymdesk SMS page (usage-based text pricing, quiet hours, opt-outs), Mailchimp guide to bot clicks (read October 2026). All conversion and time figures in the worked examples are illustrative assumptions.